/ML_SpotifyGenrePrediction

Classification machine learning algorithm for take-home interview

Primary LanguageJupyter Notebook

MachineLearning-SpotifyGenrePrediction

Machine learning classification take-home project for an interview with Thrasio, a mass Amazon FBA Buyer

Background:

  • Spotify has an internal tool that gives songs various attributes like dancibility, loudness, popularity, etc.
  • Given a dataset of songs with these attributes, we are interested in predicting the genre of the song.
  • Accuracy was the most important metric to take into consideration

In the Notebook:

  1. EDA and Feature Engineering
  2. Model training and hyper-parameter tuning
  3. Model accuracy interpretation

I utilized the Random Forest and Logistic Regression for classification, and received a max accuracy of >92%